Automating Email Sorting with Bedrock

AWS ML Blog · rss · 2026-07-09

AWS demonstrates a solution using Amazon Bedrock to automatically organize and prioritize emails, designed to help public sector agencies route large volumes of incoming mail to appropriate departments and identify urgency. The workflow routes emails to S3, triggering Bedrock model inference via EventBridge, SQS FIFO, and Step Functions. The model outputs the target department, severity, urgency, subject, and summary in a preset format. Results are written back to S3, then analyzed and visualized using Glue, Athena, and QuickSight. The article includes an example prompt and sample output tailored for UK local governments, emphasizing: - Automated routing to departments like IT, Housing, Benefits, and Waste - Prioritized handling of urgent matters - Processed data available for management dashboards and Q&A analysis

Original post →

More from Apps

Apps channel →